Distributed Database
Management Systems
Lecture 14
Minterm predicates:
given a set of simple predicates
Pr = {pr1, pr2, …, pm}for a relation R, the
set of minterm predicates M is defined
as
Virtual University of Pakistan
2
M = {m1, m2, ……, mz}
M = {mj | mj = p*k}, 1 ≤ k ≤ m, 1 ≤ j
≤z
p ∈ Pr
k
where p*k = pk or p*k = (pk)
Virtual University of Pakistan
3
PAY(title, sal)
p1: title = “El Eng”
p2: title = “Sys Ana”
p3: title = “Mech Eng”
p4: title = “Prog”
p5: SAL ≤ 20,000
p6: SAL > 20000
Virtual University of Pakistan
4
Minterm Predicates
Virtual University of Pakistan
5
m1: title = “El Eng”
m2: title = “Sys Ana”
Virtual University of Pakistan
SAL ≤ 20,000
SAL > 20,000
6
PHF-Requirements
3- minterm selectivities sel(mi):
number of tuples of the relation that
would be accessed by a user query
involving mi. Like, sel(m1) = 0 as
per data on p108
Virtual University of Pakistan
7
4- Access frequencies acc(qi):
frequency with which user
applications access data. If Q(q1,
q2, …, qn} is set of user queries,
then acc(qi) determines the
access frequency of qi within a
given period, like acc(q2) = 24
5- Access frequencies minterm
acc(mi) can also be defined from
acc(q
)
i
Virtual University of Pakistan
8
• A PHF is defined by a selection
operation on the owner
relations of the database
schema, given a relation R, its
PHF are given by
Ri = Fi
(R), 1 ≤ i ≤ w
– Where Fj is a selection formula,
which is (preferably) a minterm
predicate.
Virtual University of Pakistan
9
PHF (Simple Example)
Virtual University of Pakistan
10
PAY
Title
Sal
Elect. Eng
40000
Sys Analyst
34000
Mech. Eng
27000
Programmer
24000
Virtual University of Pakistan
11
PAY1 = σ sal ≤ 30000 (PAY)
PAY2 = σ sal > 30000 (PAY)
Virtual University of Pakistan
12
Title
Sal
Elect. Eng
40000
Sys Analyst
34000
PAY1
PAY2
Title
Sal
Mech. Eng
27000
Programmer
24000
Virtual University of Pakistan
13
• Defining simple predicates may be
tricky
– in case of continuous domain
– Balance in Fragments’ Size
Virtual University of Pakistan
14
A set of simple predicates Pr is said to
be complete if the accesses to the
tuples of the minterm fragments
defined on Pr requires
that two tuples of the same minterm
fragment have the same probability of
being accessed by the application.
Virtual University of Pakistan
15
Completeness Example
Virtual University of Pakistan
16
PROJ
pNo
pName
budget
Loc
P1
Instrumentation
3.5M
Lahore
P2
Database Dev.
2.3M
Rawalpindi
P3
CAD/CAM
1.9M
Rawalpindi
P4
Maintenance
1.6M
Peshawar
Virtual University of Pakistan
17
–Find the budgets of projects at
each location. (1)
–Find projects with budgets less
than 2M. (2)
Virtual University of Pakistan
18
• According to (1),
Pr = {LOC=“Lahore”,
LOC=“Rawalpindi”,
LOC=“Peshawar”}
• Which is not complete with respect to
(2).
Virtual University of Pakistan
19
Modify
Pr = {LOC=“Lahore”,
LOC=“Rawalpindi”, LOC=“Peshawar”,
BUDGET<=200000, BUDGET
>200000}
• Which is complete.
Virtual University of Pakistan
20
Thanks